Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Biomedical and Chemical Engineering and Sciences

First Advisor

Michael Freund

Second Advisor

Mary Sohn

Third Advisor

Yi Liao

Fourth Advisor

Anthony Smith


Transferring human abilities and knowledge to machines has been the focus of decades of work by scientists and engineers. With the progress in computation power and algorithms such as artificial neural networks, machines have surpassed humans in complicated games like chess and the Chinese game Go. Some senses can be mimicked including touch, with pressure sensors, hearing, with microphones, and vision with charged couple device (CCD)imaging chips. Yet, relatively little progress has been made with the senses of smell and taste. Machine olfaction, the ability to detect and quantify vapors in the environment, poses a range of challenges, which if overcome, could enable machines and robots to perform a vast range of tasks. In this work capacitive measurements are used to mimic olfactory receptor responses as a promising mechanism for machine olfaction. In the first chapter, the current understanding of how the human olfaction system work is described, followed by how capacitive senor technology can be employed in machine olfaction systems. After that, the current advances in different sensors technologies including their advantages and limitations will be reviewed. The technologies will be classified according to their transduction mechanism, how physical changes are converted into electrical signals that can be processed by machines. A special emphasis will be given to sensor technologies that use polymers as sensing layers. Next, the possible vapor/polymer interactions (solubility interactions) will be discussed and the models that describe the mechanism of interaction and quantify them will be reviewed. These mechanisms will be the basis of the materials chosen such that the full range of interactions are explored in the experimental results presented later. Between the different transduction mechanisms, dielectric/capacitive measurements offer the advantage of being used in miniaturized transistors that utilize the commercially available CMOS process since no post processing steps are needed. The technology is already there, in computers and smart phones. This technology can be modified to include an electronic nose into these electronic devices. In this thesis, a simple, yet effective, capacitor sensors are used for analyte sensing to give a better understanding of the dielectric/capacitive changes that happen during analyte measurements. By understanding the dielectric/capacitive changes, better and more customizable sensors can be made that can be miniaturized. The thesis has been divided into five chapters. The first chapter gives an overview of the human and machine olfaction systems. The different technologies of analyte sensing are reviewed with emphasis on the capacitive/dielectric measurements. The basics of dielectric theory is explained along with the Electrical Impedance Spectroscopy (EIS) measurements that is needed to measure the dielectric capacitive changes. Finally, the algorithm of pattern recognition used in this work to identify and classify the vapors is discussed. The second chapter presents the work on dielectric polymers as capacitive sensors. The final dielectric/capacitive signal is a combination between the solubility interactions (how much analyte goes into the polymer) and the dielectric difference (between the polymer and the analyte investigated). Dielectric, or insulator, polymers offer the advantage of having a variety of choices with different dielectric values to choose from. This gives the diversity needed by the pattern recognition algorithm to have enough information for better identifying the different vapors. EIS results gave an insight into the different process that happens inside the sensor as equivalent circuit models. Finally, the results of principle component analysis, a pattern recognition algorithm, is discussed. In the third chapter, the results of using overoxidized conducting polymers as capacitive sensors is discussed. Conducting polymers have the advantage of being easily applied into small miniaturized substrates through electropolymerization. Through copolymerization the needed diversity for different interactions and dielectric changes can be imparted. The results of capacitive changes, solubility interactions and equivalent circuits are also discussed in this chapter. In chapter four, finite element methods were used to model the behavior of chapter two and chapter three polymers’ in floating gate field metal oxide semiconductor (FGMOS) transistors. These transistors can be easily fabricated through complementary metal oxide semiconductor (CMOS) technology which is used to make transistors in electronic devices. Also, modeling of the optimum thickness to be used in the sensors is included in this chapter. Finally, chapter five is a conclusion of the work discussed in this thesis and the future potential of this work.


Copyright held by author

Included in

Chemistry Commons